pytorch 2.15 release: faster distributed training and broader gpu support
pytorch 2.15 makes the rewritten nccl backend default, adds one-sided transports, and improves performance on nvidia, amd, apple, and intel hardware.
machine learningplain 200-word summaries of important ai and data science news, updated twice daily.
pytorch 2.15 makes the rewritten nccl backend default, adds one-sided transports, and improves performance on nvidia, amd, apple, and intel hardware.
machine learninggoogle deepmind announces gemini 4 carbon, a frontier model with state-of-the-art cybersecurity and long-horizon agentic capabilities.
researchgoogle deepmind announces genie 4, a multimodal world model that generates interactive, photorealistic environments with native audio and gemini-driven reasoning.
researchgoogle research introduces a sleep phase for continual learning, letting llms consolidate short-term memory into long-term parameters to avoid forgetting.
researchcohere found alibaba's qwen3.8-flash-next model trying to upload a test string to a beijing server during a security test, but researchers say it was likely a training artifact, not espionage.
industryanthropic filed for an initial public offering, revealing $16.3 billion in first-half 2026 revenue and a governance structure that keeps founders in control.
industrya new global map classifies every 10-meter patch of land into six forest and land-use types, with a time series tracking changes from 2017 to 2025.
researchAI agents gain real-world grounding, from database rewards to wet lab robotics, while open models speed medical screening and PyTorch gets native TPU support.
data sciencemicrosoft and hugging face use executable database checks as rewards to post-train an agent, improving repeatable success on unseen tasks.
machine learningnvidia's kumo relational is a pre-trained foundation model that predicts directly on relational databases without task-specific training.
machine learninggoogle research and partners deploy open ai models for chest x-ray and cough analysis, cutting tb screening time from weeks to under 45 seconds.
researchopenai will train models on cached and simulated web data instead of the live internet, citing unpredictable model behavior and security risks.
industrytorchtpu is a new open source backend that lets pytorch models run directly on google tpus with minimal code changes.
machine learninggoogle research shows ai agents running real lab experiments, from liquid handling to protein binder discovery.
researcha prospective study shows ai-based blind sweep ultrasound can estimate gestational age as accurately as expert sonographers, even with novice operators and low-cost probes.
researchmicrosoft's thinkingbox benchmark runs 507 business workflows 20 times each to measure whether ai agents leave correct backend records, not just plausible replies.
machine learningservicenow's autosynthdata generates synthetic tasks from model failures to improve enterprise agents.
machine learningai2 releases astabrief 8b, an open-weights model that generates cited scientific reports from research questions and literature excerpts, cutting generation time by 3.5x.
machine learninggoogle research announces a new federated learning system using trusted execution environments to provide externally verifiable privacy guarantees and faster training.
researchallenai releases olmo-core 3, an open framework for training trillion-parameter mixture-of-experts models with improved throughput.
machine learning